Pigs_Can_Fly

Pigs_Can_Fly A PEDAGOGIC NOTE: Preparing a Research Proposal for a Student Research Dissertation by Clive Emmanuel & Rob Gray DRAFT 2A: November 2000 2 A PEDAGOGIC NOTE: Preparing a Research Proposal for a Student Research Dissertation Introduction Student dissertations take many different forms. These may range from, at one extreme, the full research theses expected of Doctoral (typically PhD) and Research Masters, to, at the other extreme, the usually shorter and (perhaps) more exploratory dissertations required on some undergraduate programmes. Between these two extremes lie the (usually narrowly) constrained research dissertations undertaken by students on a Taught Masters programme. In all cases, it is generally good practice to require the student to compile a research proposal. Such a proposal will probably: define the subject of investigation; make a preliminary review of the relevant literature; discuss the broad approach which the student intends to take to the investigation; and, for the more advanced student, provide some, at least initial thoughts on, research design. There are a great many texts which give guidance to students on how to approach the dissertation process and these will, on occasions, also give some advice on the preparation of the research proposal itself. (For example, see Easterby-Smith et al., 1991; Gill and Johnson, 1997; Hussey and Hussey, 1997). However, rarely do these texts actually provide an example of a research proposal which the students can use as a model. (In our experience, it is even rarer for an accounting and finance research proposal to be provided). This short paper is primarily designed to provide an illustration of a fictitious research proposal. This `research proposal’ has been developed and extensively used to help our students in this important but difficult area. However, in order that the `research proposal’ has some pedagogic value beyond simple illustration, we have found it valuable to add two brief sections. The following background provides a brief introduction that sets the proposal in context. This short paper concludes with an introductory analysis of the proposal which is intended to help the student engage critically with this – and, subsequently with their own - proposal. Background 3 A student dissertation or thesis may be undertaken for many different reasons - and these reasons may very well differ in the minds of the students and the teachers. It is probably important to attempt to be clear about which of these many possible different purposes the student dissertation is intended to serve - if for no other reason than to ensure that both staff and students have complementary (if not congruent) objectives. The dissertations we supervise have three principal objectives: to act as a training in research processes; to produce useful and interesting research results which relate to the extant accounting and finance literature; and to provide an opportunity for a student to explore further an issue of particular interest or concern to him/her. Any of our students wishing to undertake a programme of research must meet certain conditions: they must attend and pass dedicated courses in research methods; they must select an area of research which relates to existing areas of expertise and research experience amongst the supervising staff; and, except for PhD students, they must ground their intended research programme in one of the more "theoretically orientated" taught courses that they have taken and passed successfully. In addition, all dissertations are normally expected to have some empirical content. "Empirical content" in this context means that the student undertakes some interaction with research data which has either been generated for them or which they themselves collect. No restriction - other than practicality - is placed on the students' choice of research method(s), theoretical perspective or form of data analysis/interpretation. There is a basic assumption that work from research dissertations is potentially suitable to form the basis of one or more published papers. This is most obvious in the case of research degrees where publication in the substantive research journals can reasonably be expected. In the case of postgraduate taught degrees and, to a lesser extent with honours undergraduate dissertations, publication in the more accessible research journals or, perhaps, in the professional journals might be a consideration. A research proposal enters the picture at two very distinct points. The first is at application to the programme. Those applying for a place on one of the research degrees will normally be expected to have produced a research proposal of some sort. (This will generally be fairly naïve and under-developed). The second, and to our mind more important, point at which a 4 research proposal is required is within the research methods course. This latter research proposal serves four main functions in our courses: it forms part of the basis of assessment in this course; it has intrinsic pedagogic value in requiring the students to internalise and apply the material from the course; it provides the principal evidence on which the examining board decides whether or not a student can proceed to (or continue on) the research component of the degree; and, more obviously, it forms a framework for the intended research programme. Over the years in which we have taught research methods - to both postgraduates and undergraduates - it has become obvious that most students - and certainly the less able or less-focused - have real difficulty understanding what is required of a research proposal. The solution, which appears to have been successful, is to produce a fictitious research proposal that allows them to work through the elements of an intended research programme. This process of illustration seems to have complemented well the more basic advice they receive in the research methods workshops. The fictitious proposal seems to have been equally useful for undergraduate, taught and research postgraduate students. It is worth stressing, however, that different levels of proposal are, of course, required for different levels of degrees. The attached proposal would be a little too “thin” for a PhD or probably even a research masters proposal. We have pitched the proposal at the level that a good taught masters student or a good honours undergraduate might be expected to attain. The indicative, fictitious research proposal follows. 5 DRAFT RESEARCH PROPOSAL I Yam-Clueless "An examination of the Pigs-can fly hypothesis (PCFH) in a Scottish context" Introduction Whilst interest in the "pigs and flight" phenomenon dates back many centuries (Methuselah, 1952) it was only with the growth in the "Bacon School" and their development of the Pigs-can-fly-hypothesis (PCFH) (see, for example, Dog et al., 1964; Horse, 1972; Squirrel, 1984) that accounting scholars began to give serious attention to the issue (see, for example, Piggy et al., 1992). However, despite the growth in interest in the topic, Vegetarian et al. (1993) have argued that there remains considerable dispute over (i) whether the PCFH is applicable to accounting; (ii) whether the evidence can really be used to support the existence of a PCFH; and (iii) if the PCFH is accepted, at least as a maintained hypothesis, how widely applicable it is in modern accounting. This proposal relates to a proposed research project which will attempt to investigate these three points raised by Vegetarian et al (op. cit). Because of the apparent confusion over the status of the PCFH it will be necessary to undertake an extensive literature review and some theoretical analysis but the core of the dissertation will be an empirical investigation of the applicability of the PCFH in accounting in Scotland. Background My literature review on the subject is not yet complete, (See Appendix A for a list of references that I have identified as probably relevant but have not yet studied in any detail). As a result this section of the proposal is very sketchy. However, from a reading of Danish et. al. (1993) and Rindless (forthcoming) it seems that we may identify the following principal themes. Danish et al. (1993) define the PCFH as relating to the phenomenon that unproductive assets and badly designed products may still produce positive earnings streams in a business if (a) the joint costs of the assets and (b) the marketing expenditure on the products are treated in an optimistic way thus permitting previously "un-economic" business activities to become "economic". Rindless (forthcoming) then relates this to accounting by arguing that optimistic approaches to costing and marketing challenge the basic prudence concept. Further, if the PCFH is a reality, accounting bodies must consider permitting a reduction in, especially (i) depreciation charged against such assets (ii) ignoring the lower of cost and NRV rule for inventory and (iii) capitalization of marketing expenditure. In these circumstances, they argue, companies may thus be able to exploit marginal returns available on previously loss-making activities. However, the very basis of the PCFH has been challenged. Smoky (1989), for example, argues that not only is there little evidence for the hypothesis but also reasoning behind the hypothesis counters the principles of both economic reality and generally acceptable accounting principles. He 6 goes on to argue that the employment of the PCFH will send misleading signals to investors and runs counter to the principles of reliable stewardship. The evidence is, indeed, patchy. Piggy et al (1992) derive the major accounting application of the PCFH by examining reinvestment data in 1000 failed firms in the Agricultural livestock industry in Mexico in 1946 and then test the thesis against current USA decisions taken in the engineering and foods industries in the 1980s'. They argue that (a) marginal productivity decisions could have been improved by application of the PCFH and (b) the "dumping" of "pigs" on developing market economies can be shown to be economically desirable. Vegetarian et al (1993) by contrast could find no evidence for the hypothesis in their own research in the Icelandic Rock Music Industry. Methodology The central issues to be resolved before any investigation of the PCFH can be undertaken are: (i) what do we mean by "pigs" (ii) what do mean by "flight" (iii) what theoretical perspective we will take on the issues. The principal conclusions of Vegetarian (1994) centre on the contention that the Bacon School has treated as "scientific" a series of concepts which are, in essence, "social constructionist". That is, there is an ontological misconception over the thing being studied. To therefore employ the functionalist statistical analysis so popular with the Bacon School is to commit, what Vegetarian calls, "epistemological hari kari". Vegetarian goes on to argue that "pigs flying" is a belief issue not an economic issue. Oddly enough, Cat (1990), from a radical critical perspective, comes to a similar conclusion. However, Cat (1990) goes on to argue that even if pigs can fly they can only do so by means of exploitation of customers, especially in the developing world. Of particular note, Sty and Swill (1987) reach the same conclusion as Cat by demonstrating that the PCFH is an "economically rational" strategy for any manager in an agency framework. The same conclusions (although significantly different moral interpretations) reached from such different political points of view suggest that the conclusion may bear investigation. (We should note that Vegetarian (1994) is sceptical of the analysis undertaken by Sty and Swill for the social constructionist reasons given above). There are, therefore, a wide range of theoretical perspectives which can be successfully applied to the PCFH. Central to them is the choice of ontology and the consequential epistemological implications. This proposal will follow Smoky (1989) in her suggestions that the existence or otherwise of "pigs" and "flight" can be considered irrelevant if one focuses on managers' and accountants' perceptions. As a result, it is the actors' beliefs about PCFH that matter, not the hypothesis itself. If this is then considered within, as Smoky suggests, a control theory framework (Somebody, 1973; 1975), it may be possible to examine the extent to which managers and accountants are motivated to belief or unbelief, why they perceive the belief as desirable and the extent to which the belief leads to self-fulfilling "reality". Smoky has also pointed out that there would appear to be a temporal and cultural element in 7 the PCFH which suggests that time-and-country-specific analysis are more likely to yield useable results. Method The research will comprise the following elements: a) an extensive literature review. This will first of all attempt an analysis of the different ontological positions taken in the literature and, second, assess the evidence in relation to those positions. (This will follow closely the work of Smoky and Vegetarian); b) a review of the control theory framework (Somebody, 1973) and explanation of the PCFH within this framework; c) a series of interviews with financial directors in Scottish companies in the Haggis Farming Sector. (Danish et al., and Smoky both emphasise the industry-specific nature of the PCFH and I have now made contact with 4 local companies). The interviews will be semi-structured (See Appendix B for the proposed interview format) and will explore, generally, practitioners' view of the PCFH. d) a postal questionnaire to all Finance Directors in the Haggis Farming Industry. This will be short and based on the experience in the interviews. It is anticipated that the principal focus will be to elicit respondents' beliefs about the existence of the PCFH, the extent to which they have "pigs", whether they have tried to fly pigs, what outcomes they perceive have resulted from this and, most importantly, what accounting procedures have been employed. I intend to try and produce a pigs-score (Swill, 1984) on the basis of these (Appendix C give more details). e) if time permitted, I would also like to compare the questionnaire answers with industry data (available on Haggis-stream) to try and assess whether there is any evidence of improved market performance from companies with high pig-scores. This might best be analysed by (a) tests of differences and (b) longitudinal analysis of high pig-score companies. Research Design The choice of theoretical framing, research methods and sampling frame itemised above is not arbitrary. As Fussy and Particular (1999) so persuasively argue, whilst research design inevitably reflects the individual researcher’s experiences, methodological preferences and personal interests, this is not an excuse for an arbitrary approach to seeking to answer the research question. “The researcher must seek to justify the research choices made. This is not just so that the choices made are explicit, but also to establish that the approach chosen is the one most likely to lead to the best available answer to the research question – given all the research’s preferences and constraints” (op cit. p34). Any research design choice is a strategic one, (Fussy and Particular, 1999, p23). In my research, the principal choices made are (a) to follow the work of Smoky and Vegetarian and (b) to frame the work within control theory. This choice is made to avoid either the specifically functional 8 analyses offered by the Bacon school – and which have offered conflicting results – as well as seeking to avoid the extremes of the ideological debate as represented by Sty and Swill and by Cat. Thus whilst aspects of the PCFH can be treated as structured problems, it is my judgement that the problem itself is better considered as an unstructured problem which is better addressed through a flexible and exploratory study – albeit one which draws from structured methods where possible but which employs a high degree of triangulation. Consequently, the choice of semi-structured interviews as the initial research method seems highly appropriate as this will allow for exploratory discussion which is not necessarily pre-determined by prior theorising. (The obvious alternative, grounded theory, was not felt to be appropriate given the intentions of the later stages of the project). The choice of the Haggis Farming Sector is, I am afraid, largely arbitrary and reflects: my interests; local importance; and where I could obtain access. Nevertheless, if the industry-specificity of the PCFH is valid, then a detailed examination of a single industry might help us better understand why that is. (I cite again, Danish et al., and Smoky, in my defence). The choice of a postal questionnaire to follow up from the interviews was also an inevitable choice given the earlier decisions. That is, given the ontological doubt surrounding the PCFH and the social constructionist arguments reviewed above, an attempt is being made here to both capture attitude and to seek to establish whether there is any industry generalisability in the hypothesis. In my understanding of the arguments of methodology, only if the generalisability of the hypothesis in a single setting can be shown, would more analytical, but strictly functionalist testing be justified. Conclusion It is explicitly recognised that this is an exploratory study and so, whilst hypothesis testing may be possible in part (e) above, the bulk of the dissertation will be concerned with trying to understand and isolate the putative PCFH. I expect to find that respondents who believe in it do act to make it happen but that only some of these are "successful". It is unlikely that this dissertation will be able to clearly isolate "success". As a result the dissertation will be very limited in its conclusions. There are also very notable omissions - perhaps the most important relate to financial and other stakeholders whose perceptions and views are argued to be essential to the PCFH (see, for example, Blandings and Castle, 1994). 9 Discussion of the Proposal The proposal follows the conventional pattern of addressing: 1. the research question; 2. a review of the literature; 3. which helps to define both the research question and the terms used in that question; 4. the methodological assumptions employed in prior research and which will be applied here; 5. the research methods to be used (together with a simple justification); 6. the potential findings and where their contribution to the literature might lie; 7. limitations; and 8. conclusions This is a useful template for any research project proposal but it is especially important to note that, especially in the early stages of the project, items 1-5 need to be regularly re-visited, examined and iterated. For example, determination of the literature relevant to the research question – and, especially, to the way in which the question is posed – will typically be of critical importance and is often time-consuming. Reconstruction of the literature – both that within accounting and finance and that outside the discipline – should help clarify which prior research can be considered as relevant to the research problem (Ryan, Scapens and Theobald, 1992). Having an efficient referencing system which, from the start, catalogues journal articles, books and other sources, using key words is virtually essential in this context. As the research question becomes more transparent, previously read material may change its status from (say) peripheral to central (or vice versa). Equally, the proposal properly addresses methodological assumptions such as ontology, epistemology and human nature (Burrell and Morgan, 1979). These are difficult concepts for any researcher but typically prove very difficult indeed for research students to grasp. This is not just because of the terminology but, more importantly, because the concepts require an, 10 often unfamiliar, degree of introspection from the researcher. The choice of a more positivist or subjectivist (or phenomenological, see Morgan and Smircich, 1980) stance is likely to be influenced by the researcher’s bias, skills, interests and prior experience. A positivist/functionalist position believes reality can be independently observed whereas a subjectivist position views reality, (expressed simply), as conveyed by meanings individuals and groups associate with events or phenomena. The correct combination of underlying methodological assumptions is that which matches, not only the predisposition of the researcher but also, the way in which the research question is phrased. The assumptions may very well be implicit within the research question but the relative emphasis on “why?’, “what?”, “when?”, “how?” etc.. may well indicate which assumptions are central to the question and the form in which its `answer’ can be expressed. The PCFH proposal raises the methodological question of whether ontology precedes epistemology. That is, can the researcher’s beliefs about the way in which s/he observes `reality’ come before assumptions about what is regarded as `knowledge’? It is difficult – perhaps even impossible – to be categorical in a general sense about this issue. However, if there is a constant iteration between the research question, the methodological assumptions and the chosen research methods it becomes almost impossible to state that ontology informs epistemology or vice versa. Uncomfortable as it might seem, the constant iteration means that what is defined as reality by the research is informed by what the researcher regards as relevant – or, even, persuasive - knowledge. Of course, those wedded to a grounded theory approach, may well argue that ontology is paramount (see, for example, Glaser and Strauss, ; Strauss and Corbin, ). However, students on taught programmes and/or with greater pressure of time thereby restricting their capacity for sophisticated refection, the safer route is usually to give epistemology the priority where a relevant literature provides a foundation for the proposal. Where, however, there is no acceptable theory, or (worse) confounding theories, the researcher has to potentially make more sophisticated choices. Grounded theory or `middle-range thinking’ (Laughlin, 1995) offer two possible ways through this maze. 11 The proposal also highlights the way in which the research question is specified. In essence, the researcher has moved from a broad and wide-ranging appreciation of the literature to an increasingly specific point of focus. This might be thought of as a pyramid, with a broad base of exploratory reading from different perspectives and different disciplines. Careful reconstruction of that broad base through an examination of the epistemological and knowledge claims of that literature has led to a focus. This is crystallised through (a) a homing in on the (what is called here) the control theory framework and (b) a decision to commit to a specific theme within the broad literature. Together, these processes bring the researcher to a point where a focused piece of research is a possibility. Up until this point, the researcher has relatively little to show for all the hours spent on reviewing and introspection. Once the focus of the pyramid begins to emerge, the choices of operationalising the enquiry can be addressed. The PCFH illustration calls for triangulation of research methods. Interviews are to be followed by a principally analytic questionnaire as the means of data collection. The choices faced by the researcher are as follows. First, is it necessary to collect original data or would, for example, a meta-analysis or the use of third-party collected data source (e.g. share price data) be sufficient? Second, what is the appropriate sequence of research methods? Third, what kind of data will be collected by these methods and how will it be systematically and rigorously analysed? None of these are trivial questions. We can really only answer them by reference to the way in which the research question is posed and the reliance we place upon (and choices we make about) the prior literature reviewed. at least in terms of One place where the proposal may actually begin to break down – its methodological coherence – is in the (implicit) assumptions underlying the form of interviews and questionnaires chosen. That is, the choice of a semi-structure – as opposed to an unstructured – form of interview suggests that the researcher has taken some justification from the prior literature about the degree to which the research question is already reasonably well focused. Equally, the use of a questionnaire – particularly one which would appear from the proposal to be fairly closely focused (i.e. rather more analytic as opposed to exploratory or descriptive) - suggests an increasingly structured enquiry. If, indeed, the researcher can justify the research question and its link to both methodology and method, then the question arises 12 whether or not the research would be bettered designed around an analytic, structured questionnaire or experiment. In this latter case a more formal specification of independent and dependent variables can be achieved, issues of construct validity can be assessed and the data collection becomes (apparently) more straightforward and positivist. The proposal has chosen to take an initially subjectivist approach which iterates towards a more positivist stance. There would appear from the proposal to be no reason why the opposite approach might not be taken. That is, an approach which started from an essentially positivist approach to triangulation and then fed back the statistically significant results to focus groups, fora in the company or via interactive interviews. Again, no immutable rules can be offered but the strands in the research can be evaluated in this way to assess the degree to which the research has been thoughtfully designed and whether, as in this case, the opportunity for replication exists. The projected findings of the PCFH proposal are important because they demonstrate the extent to which the original research question may be answered. Every piece of worthwhile research has its limitations but it is up to the researcher to demonstrate that, on balance, the right (or at least, the best available) choices relative to the research question have been made. Elucidating these limitations provides opportunities for future research and researchers, thus honesty and clarity are important considerations in describing the research process. Concluding Remarks Is the research process, as it must seem to so many students new to research, simply one of constant doubt, iteration and blood, sweat and tears? Certainly not, as a colleague Peter Armstrong was wont to say – if you are not having fun and trying to make a difference, don’t bother. Worthwhile research requires the researcher’s commitment and interest (even passion); it requires a careful introspection and analysis of his or her skills; and it requires that the bias that we all display to be formally recognised. We are all different in these respects and by appreciating these differences in as unblinkered a fashion as possible, we should be able to 13 appreciate the research which colleagues and students undertake – whether or not it lies outside our own accepted methodological assumptions, methods or perspectives.